American Journal of Transplantation 2015; 15: 2261–2264 Wiley Periodicals Inc.
Copyright 2015 The American Society of Transplantation and the American Society of Transplant Surgeons doi: 10.1111/ajt.13243
Percutaneous Retroperitoneal Splenorenal Shunt for Symptomatic Portal Vein Thrombosis After Liver Transplantation C. Pulitano1,2, C. Rogan3, C. Sandroussi1, D. Verran1, G. W. McCaughan1,2, R. Waugh3 and M. Crawford1,* 1
Australian National Liver Transplant Unit, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia 2 Centenary Research Institute, University of Sydney, Camperdown, New South Wales, Australia 3 Radiology Department, Royal Prince Alfred Hospital, Sydney, New South Wales, Australia Corresponding author: Michael Crawford, Michael.Crawford(RPAH)@sswahs.nsw.gov.au
Acute or recurrent bleeding from ectopic varices is a potentially life-threatening condition in rare patients with extrahepatic complete portal vein thrombosis (PVT) after liver transplantation (LT). In this setting, the role of interventional radiology is very limited and surgical shunts, in particular splenorenal shunts are usually used, despite the high associated mortality. We present the first reports of the clinical use of a new minimally invasive technique, percutaneous retroperitoneal splenorenal shunt (PRESS), in two LT recipients with life-threatening variceal hemorrhage secondary to PVT. Both patients had a successful PRESS using a transplenic approach with resolution of bleeding, avoiding the need for a potentially complicated laparotomy. The PRESS procedure is a useful addition to the interventional armamentarium that can be used in cases unsuitable for surgical shunt, and refractory to endoscopic management. In the future, this technique may be an alternative to surgical shunts as the standard procedure in patients with extra-hepatic PVT, just as the transjugular intrahepatic portosystemic shunt (TIPS) procedure has become for the management of portal hypertension in the absence of PVT. Longerterm follow-up will be needed to establish the longterm success of this procedure. Abbreviations: IVC, inferior vena cava; LT, liver transplantation; PRESS, percutaneous retroperitoneal splenorenal shunt; PV, portal vein; PVT, portal vein thrombosis; TIPS, transjugular intrahepatic portosystemic shunt Received 23 September 2014, revised 25 January 2015 and accepted for publication 25 January 2015
Introduction Bleeding gastrointestinal varices are rare and potentially life-threatening conditions (1). They are usually found in patients with portal hypertension and a history of gastrointestinal surgery. Typically, they are managed by radiological interventions such as transjugular intrahepatic portosystemic shunt (TIPS) and transhepatic embolization (2). However, in patients with portal vein thrombosis (PVT), an alternative approach is necessary. In patients with PVT after liver transplantation (LT) with preserved liver function, distal splenorenal shunt has been used for the management of refractory variceal bleeding (3). However, it requires a major open operation and it is associated with significant morbidity and mortality. Survival of patients who have undergone surgery is dependent on liver function but the mortality can be as high as 80% (2). We report two cases of gastrointestinal bleeding associated with PVT after LT, where surgical shunting was relatively contraindicated, which were successfully managed by a new minimally invasive technique, percutaneous retroperitoneal splenorenal shunt (PRESS) (4).
Case Reports Case 1 A 23-year-old man presented with hematochezia and anemia. He had a history of cystic fibrosis–related liver disease, chronic pancreatitis and pancreatic insufficiency. He had previously undergone a small bowel resection. He underwent LT at the age of 17, which was complicated 5 years later by PVT presenting with recurrent lower gastrointestinal hemorrhage and thrombocytopenia in the presence of normal graft function. Colonoscopy showed portal hypertensive colopathy without evidence of recent bleeding. Visceral computed tomography (CT) angiography demonstrated chronic PVT with cavernous transformation, the presence of severe portal hypertension with extensive varices and a heavily calcified pancreas. The patient continued to experience ongoing bleeding per rectum with hemodynamic instability requiring ICU admission and inotrope infusion. Six units of packed red cells were also transfused. Surgery was considered to pose an extremely 2261
Pulitano et al
Figure 1: Preinterventional CT scan Case 1 (A) and Case 2 (B). It is evident the close relationship between the splenic (gray arrow) and the left renal vein (black arrow).
high risk because of the underlying cystic fibrosis, previous abdominal surgery and chronic pancreatitis in the context of severe portal hypertension. Considering the close relationship between the splenic vein and left renal vein evident on CT scan (Figure 1), a radiological approach was devised. The procedure was performed under general anesthesia. Initial angiography via the celiac and superior mesenteric arteries failed to demonstrate a bleeding point. Planning a transplenic approach, the left renal vein was first cannulated using the femoral vein as access. The process for splenic access was as follows: A 19 G thinwall 15 cm Cook Medical (Bloomington, IN) trochar needle was advanced under ultrasound guidance into a branch splenic vein. A benston guidewire was advanced into the splenic vein. A 6 Fr sheath was inserted and a JB1 catheter advanced into the splenic vein and check venography performed. A 12 Fr 13 cm checkflow sheath was inserted into the splenic vein. Through this, the 9 Fr angled tip checkflow sheath (from the TIPS set) was inserted over the wire and its tip directed inferiorly toward the left renal vein. Using a rotating C-arm to facilitate positioning, the TIPS needle was advanced from the splenic vein across retro-peritoneal tissue and into the left renal vein. A Viatorr PTFE graft 10 60 mm covered stent (WL Gore and Associates, Flagstaff, AZ) was deployed from the distal splenic vein into the inferior vena cava (IVC). There was good flow present on both the completion splenic angiogram and the injection through the venous sheath. Finally, the tract was closed with a multitude of small pellets of gelfoam and a 2cc injection of Histoacryl and Lipiodol (Figure 2). Final completion splenic angiogram showed no extravasation from the spleen. A check CT scan 24 h later showed occlusion of the stent, so the patient returned to angiography. The occluded stent at the distal end was protruding some way into the IVC. A Shepherd hook catheter was passed via the femoral vein, 2262
and a guide wire was easily passed across the soft clot in the occluded stent into the splenic vein. A 12 60 mm Zilver stent (Cook Medical) was deployed with the top end in the main splenic vein. There was still insufficient flow and, therefore, a second Zilver stent 12 4 mm was deployed flexed at the lower end and definitely into the renal vein. After both stent deployments, a 12 mm balloon was used to fully dilate the stents (Figure 3). After the second procedure, hematochezia ceased, suggesting successful hemostasis. There was no development of splenic hematoma or hemoperitoneum on a CT 4 days later. The patient was discharged 6 days after the procedure without further complication. At 18 months follow-up, the shunt remained patent on CT scan (Figure 4). The patient has had no further complications related to portal hypertension and has had no clinically apparent encephalopathy. Case 2 A 29-year-old man presented with melena and anemia. He had a history of ulcerative colitis, primary sclerosing cholangitis and diabetes. The patient had also permanent
Figure 2: Embolized splenic tract.
American Journal of Transplantation 2015; 15: 2261–2264
Percutaneous Splenorenal Shunt
extravization of contrast to suggest a bleeding point. The patent splenic vein and the left renal vein were again shown to be in close proximity (Figure 1B).
Figure 3: Transsplenic catheter angiography with the catheter tip in the renal vein shows the patent stent.
cognitive deficit as consequence of severe head injury. He underwent LT with hepaticjejunostomy at the age of 11 and was retransplanted at the age of 18 because of chronic rejection. He had hyperbilirubinemia, and cholestatic liver enzymes, but other synthetic functions were normal. Despite ongoing melena, upper endoscopy and colonoscopy failed to demonstrate the source of bleeding. Due to instability, the patient was transferred to ICU where continued to experience ongoing bleeding requiring the transfusion of 4 units of packed red cells, and increasing amounts of noradrenalin infusion. A CT mesenteric angiogram showed PVT, and diffuse large venous collaterals around the duodenum and pancreas, but no significant gastric or esophageal varices, and no
The patient was deemed a moderate to high risk for surgical shunt formation and, therefore, recommended for the PRESS procedure. A femoral venous sheath was placed and a wire and catheter were passed into the left renal vein. Using ultrasound guidance an 18 gauge thinwall (Cook Medical), trochar needle was inserted through the spleen into a first order splenic vein. An attempt was made to pass a TIPS needle from the splenic to the renal vein, but it was unsuccessful. The Ross modified Colapinto needle was then advanced superiorly from the left renal vein. The Benston wire was then snared from within the splenic vein with an EnSnare 20 mm endovascular snare system (Merit Medical, Salt Lake City, UT) used to establish a through and through wire. Balloon dilatation of the track was performed using a 12 mm balloon. Two 12 mm self-expanding stents uncovered stents were placed, then a covered Viabahn PTFE stent 12 mm 80 (WL Gore and Associates) was deployed inside these two stents to cover the extravascular tract. Splenic angiography showed all contrast entering the shunt. A large amount Gelfoam pellets were injected in the spleen. An abdominal duplex performed the day after the procedure confirmed that the shunt was patent. The patent shunt allowed for a confident endoscopic approach to his biliary tree. Six days after the PRESS procedure, he underwent a single balloon enteroscopy for a concomitant stenosis of hepaticojejunostomy anastomosis. The patient had no further bleeding episodes and was discharged 6 days later. At 6 months follow-up, duplex imaging confirmed the patency of the shunt and no further bleeding episodes.
Figure 4: Follow-up CT scan at 12 months shows patent stent.
American Journal of Transplantation 2015; 15: 2261–2264
Portal vein stenosis and PVT are rare late complications of LT with an incidence of less than 2% (5). Severe stenosis or PVT usually results in portal hypertension with the consequent risk of variceal bleeding. The role of percutaneous treatment for acute or recurrent variceal bleeding in patients with PVT has been very limited. TIPS for patients with PVT and cavernous transformation has been proposed, but the real utility remains controversial (6). To date, small case series of percutaneous embolization of varices are the only percutaneous procedures reported in the literature (4,7,8). In patients with recurrent variceal bleeding, surgical shunts have been reported successfully in LT recipients with PVT, and would be considered the standard of care (5). Mesocaval shunts are usually avoided in LT because potential risk of thrombosis of the superior mesenteric vein and subsequent loss of the hepatic allograft. In the distal splenorenal shunt, the risk of jeopardizing a possible retransplant is reduced because the dissection is away from the hilum (3). As a 2263
Pulitano et al
consequence, distal splenorenal shunt has been the most commonly used shunt for patients with symptomatic PVT after liver transplant (5). The idea of a nonsurgical alternative to distal splenorenal shunt is based on the high invasiveness and complication rate associated with open redo surgery, and the favorable venous anatomy. The retroperitoneal cavity is ideal for a percutaneous vascular shunt because in humans the left renal vein is relatively close to splenic vein, particulary in patients with severe portal hypertension and consequent splenomegaly (Figure 1). Furthermore, the perirenal space is constrained by the fascia of Gerota, which limits any perirenal bleeding. The technical feasibility of a PRESS was demonstrated by Kaminou et al in 5 swine in 1998 (9). Following this initial description, no other author has published the use of this technique. To our knowledge, this is the first report of the clinical use a modified PRESS technique. An important difference between our technique and the one described by Kaminou et al is that we accessed the splenic vein through the spleen and not transhepatically. A transhepatic approach would not be feasible in the presence of PVT as in our two patients. A percutaneous transsplenic approach has been previously described to access the PV (4,7,8). However, because of the fragility of splenic parenchyma, percutaneous transsplenic PV interventions have been rarely used in clinical practice. Embolization of splenic percutaneous puncture tracts with gelatin sponge or Amplatzer Vascular Plug is well-described techniques after splenic biopsy and access with low risk of subsequent splenic hemorrhage (7,10). The two patients described in this report had life-threatening hemorrhage, with only one alternative treatment option, the surgical creation of a surgical shunt. Nevertheless, this is a major operation with high risk of significant morbidity and mortality in patients with poor general condition or hepatic reserve (2). There are reports of mesorex, mesocaval and splenorenal shunts for complicated late PVT after LT, each leading to successful splanchnic venous decompression without compromising liver function (11,12). We elected to use the PRESS procedure in our first patient when the surgical risks of operating in a hostile abdomen were thought high enough to justify the unknown risks of the procedure. The second patient had a significant chance of having a similarly hostile abdomen, and we were aware of the success of our first procedure. Despite the first case being complicated by primary failure followed by secondary patency, we have shown the feasibility of this technique in clinical practice.
In conclusion, this is the first description of the successful use of a modified PRESS technique in humans. In selected LT recipients with normal liver function and PVT, in the face of life-threatening GI bleed, PRESS is a useful addition to the surgeon’s armamentarium. Indications for PRESS might be expanded into the non–liver transplant population where surgical intervention is risky, and in the future it may obviate the need for surgical spleno-renal shunts entirely.
Disclosure The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.
References 1. Khanna R, Sarin SK. Non-cirrhotic portal hypertension—Diagnosis and management. J Hepatol 2014; 60: 421–441. 2. Biecker E. Portal hypertension and gastrointestinal bleeding: Diagnosis, prevention and management. World J Gastroenterol 2013; 19: 5035. 3. Marino IR, Esquivel CO, Zajko AB, et al. Distal splenorenal shunt for portal vein thrombosis after liver transplantation. Am J Gastroenterol 1989; 84: 67–70. 4. Lee JY, Song S-Y, Kim J, et al. Percutaneous transsplenic embolization of jejunal varices in a patient with liver cirrhosis: A case report. Abdom Imaging 2012; 38: 52–55. 5. Porrett PM, Hsu J, Shaked A. Late surgical complications following liver transplantation. Liver Transpl 2009; 15: S12–S18. 6. Senzolo M, Tibbals J, Cholongitas E, Triantos CK, Burroughs AK, Patch D. Transjugular intrahepatic portosystemic shunt for portal vein thrombosis with and without cavernous transformation. Aliment Pharmacol Ther. Blackwell Publishing Ltd 2006; 23: 767– 775. 7. Zhu K, Meng X, Bin Z, et al. Percutaneous transsplenic portal vein catheterization: Technical procedures, safety, and clinical applications. J Vasc Interv Radiol 2013; 24: 518–527. 8. Tuite DJ, Rehman J, Davies MH, Patel JV, Nicholson AA, Kessel DO. Percutaneous transsplenic access in the management of bleeding varices from chronic portal vein thrombosis. J Vasc Interv Radiol 2007; 18: 1571–1575. €sch J, Yamada R, et al. Percutaneous retroperitoneal 9. Kaminou T, Ro splenorenal shunt: An experimental study in swine. Radiology 1998; 206: 799–802. 10. Dollinger M, Goessmann H, Mueller-Wille R, Wohlgemuth W, Stroszczynski C, Heiss P. Percutaneous transhepatic and transsplenic portal vein access: Embolization of the puncture tract using €ntgenstr 2014; 186: 142– amplatzer vascular plugs. Fortschr Ro 150. 11. Tepetes K, Tzakis A, Tzoracoleftherakis E, Starzl T. Portosystemic shunt for the treatment of portal vein thrombosis following orthotopic liver transplantation. Transpl Int 1994; 7: S117–S118. 12. Krebs-Schmitt D, Briem-Richter A, Grabhorn E, et al. Effectiveness of Rex shunt in children with portal hypertension following liver transplantation or with primary portal hypertension. Pediatr Transplant 2009; 13: 540–544.
American Journal of Transplantation 2015; 15: 2261–2264